How do designers generate new ideas? Design heuristics across two disciplines Seda Yilmaz1, Shanna R. Daly2, Colleen M. Seifert3 and Richard Gonzalez3 1 Department of Industrial Design, Iowa State University, 158 College of Design, Ames, IA 50011, USA 2 Department of Mechanical Engineering, 2350 Hayward, University of Michigan, Ann Arbor, MI 48109-2125, USA 3 Department of Psychology, 530 Church Street, University of Michigan, Ann Arbor, MI 48109-1043, USA Abstract Research supports the central role cognitive strategies can play in successful concept generation by individual designers. Design heuristics have been shown to facilitate the creation of new design concepts in the early, conceptual stage of the design process, as well as throughout the development of ideas. However, we know relatively little about their use in differing disciplines. This study examined evidence of design heuristic use in a protocol study with 12 mechanical engineers and 12 industrial designers who worked individually to develop multiple concepts. The open-ended design problem was for a novel product, and the designers' sketches and comments were recorded as they worked on the problem for 25 min and in a retrospective interview. The results showed frequent use of design heuristics in both disciplines and a significant relationship to the rated creativity of the concepts. Though industrial designers used more heuristics in their concepts, there was a high degree of similarity in heuristic use. Some differences between design disciplines were observed in the choice of design heuristics, where industrial designers showed a greater emphasis on user experience, environmental contexts, and added features. These findings demonstrate the prevalence of design heuristics in individual concept generation and their Received 9 December 2014 Revised 29 July 2015 effectiveness in generating creative concepts, across two design domains. Accepted 26 August 2015 Published Key words: creativity, design heuristics, cognitive strategies, engineering design, industrial design Corresponding author C. M. Seifert [email protected] Published by Cambridge 1. Introduction University Press c The Author(s) 2015 How do designers create new concepts? When faced with the task of generating Distributed as Open Access under a CC-BY 4.0 license a new design for a novel or existing product, designers find a wide variety of (http://creativecommons.org/ ways to think of new ideas. The ability to take a problem and generate multiple, licenses/by/4.0/) varied solutions that can lead to new, creative outcomes is often referred to as Des. Sci., vol. 1, e4 the concept generation or ideation stage of design (Simon 1969). Finke, Ward & journals.cambridge.org/dsj DOI: 10.1017/dsj.2015.4 Smith (1992) divided ideation strategies into generative (e.g., analogical transfer and association) (Lloyd & Scott 1994) and exploratory (e.g., context shifting and hypothesis testing). Perhaps the most ubiquitous method is called `brainstorming', a group process that involves suspending evaluation and generating as many 1/29 Downloaded from https://www.cambridge.org/core. IP address: 170.106.202.58, on 02 Oct 2021 at 03:42:47, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/dsj.2015.4 different ideas as possible (Osborn 1957). Variants of this method, including `brainwriting' (Paulus & Yang 2000), focus on developing a large quantity of ideas. The ideation stage promotes creativity when many different designs are generated for later evaluation (Christiaans & Dorst 1992). Fluency has long been used as a measure of creative ability (Clark & Mirels 1970; Torrance 1972), and generating more ideas will logically allow the selection of a better idea from the set. But at times, continuing to generate a diverse set of ideas may prove challenging. One reason is that designers can become `fixated' (Jansson & Smith 1991; Purcell & Gero 1996), where their attention is focused on a single past example or on one new idea. As a result, other possible designs are never generated, limiting the diversity and therefore the quality of ideas generated, and opportunities for novelty and innovation are missed (Ullman, Dietterich & Stauffer 1988; Ball, Evans & Dennis 1994). To assist designers in ideation, a variety of ideation tools exist (c.f. Fogler & Le Blanc 1995). These range from more complex approaches like morphological analysis (Pahl & Beitz 1996) to simple checklists (Osborn 1957). `Concept mapping' helps by identifying links between concepts relevant to a design (Plotnick 1997). Synectics (Gordon 1961) provides a series of `idea triggers' to help designers generate new concepts. The SCAMPER technique (Eberle 1995) guides with questions such as, `What else can this be used for?' Another tool based in engineering, the theory of inventive problem solving (TRIZ) (Altshuller 1984), provides a system to modify solutions by solving contradictions or tradeoffs based on past product patents. Designers naturally generate ideas also without tools (Purcell & Gero 1996); these natural approaches are developed based on designers' experiences and preferences for problem solving (Kirton 2004). However, it can be difficult for designers to describe their own cognitive thought processes (Daly et al. 2010), which may occur largely unconsciously (Nisbett & Wilson 1977). For this reason, think-aloud protocols (Ericsson & Simon 1993) are used to study designers as they speak their thoughts while working through a design problem. Protocols have been shown to be effective in understanding designers' thoughts without interfering with their natural thinking processes (Atman & Bursic 1998). Many empirical studies have investigated the cognitive processes of individuals during the design process (Adams & Atman 1999; Christiaans & Dorst 1992; Dorst & Cross 2001; Stauffer & Ullman 1988; Dahl & Moreau 2002; Kavakli & Gero 2002; Okudan, Ogot & Shirwaiker 2006; Hernandez, Shah & Smith 2010; Lawson 1980). Important features about the cognitive processes within design have been discovered, such as the observation of unexpected discoveries (Suwa, Gero & Purcell 2000), `novel design discoveries' (Akin & Lin 1995), and opportunism in design (Guindon 1990; Ball & Ormerod 1995). This paper gives a cognitive account of generalized patterns observed in designers' concepts. The outcomes of a think-aloud protocol study are presented to explore how design heuristics are used by designers from two disciplines of conceptual design: mechanical engineering and industrial design. We also investigated (1) how designers used design heuristics in generating multiple candidate designs, (2) how the use of heuristics impacted the variation among concepts and the creativity of concepts generated by designers, and (3) how the two groups of designers differed, if at all, in their use of design heuristics. 2/29 Downloaded from https://www.cambridge.org/core. IP address: 170.106.202.58, on 02 Oct 2021 at 03:42:47, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/dsj.2015.4 Figure 1. Novague, a Prague design studio, has proposed a rocking-chair concept that converts rocking motion to power a reading light (from http:// www.tuvie.com/novague-rocking-chair-generates-energy-to-light-a-led-lamp/). 1.1. Design heuristics `Design heuristics' are defined as cognitive `shortcuts' that point toward useful design patterns (Daly et al. 2012c; Yilmaz, Seifert & Gonzalez 2010; Yilmaz & Seifert 2011). `Heuristic' commonly refers to strategies that use readily accessible information to control problem-solving processes (Pearl 1984), and has been shown to be advantageous in applied problems (Gilli, Maringer & Winker 2008; Goslar 1993). Cognitive heuristics help individuals arrive at a solution (Nisbett & Ross 1980) through readily accessible information in memory in order to control and direct problem-solving processes. Importantly, cognitive heuristics are not guaranteed to produce a successful solution, but can help to quickly identify possible solutions. Behavioral research in domains like firefighting found that experts use cognitive heuristics automatically when facing new problems, and that heuristics can lead to fast, effective solutions (Klein 1998). Designers have also been shown to follow trains of thought to arrive at partial solutions with little cognitive cost (Guindon 1990). For example, a designer may display opportunism (Ball & Ormerod 1995) by combining the energy source for two different functions – reading and rocking – into a rocking chair that powers a reading light (see Figure1). Once this concept has occurred to the designer, does she then look for other instances of the same type of opportunity? From this experience, the designer may learn that opportunities may come in the form of two functions that occur together, and so afford the possibility of merging their energy source (Yilmaz & Seifert 2010). This conclusion goes beyond observing cognitive processes at a general level, such as noting that an opportunity occurred; instead, the specific type of idea generated – `use the same energy source' – may be useful to try again when functions co-occur. Design heuristics are specific content patterns reflecting the cognitive strategies used to create new concepts (Daly et al. 2012c; Yilmaz & Seifert 2011, 2010). In one study, an experienced industrial designer working on a universal access bathroom generated a concept where the sink, toilet, and shower